lawfulcactus

Joined 7 years ago

Karma
53
Submissions
24
Member since
Apr 25
[ my public key: https://keybase.io/lawabidingcactus; my proof: https://keybase.io/lawabidingcactus/sigs/1lvRvLit7ABIq-tFsrzUjGL-8SWHzBecaWpLwMHfUqE ]

PGP fingerprint: 09396E67CCFD945008A4C4B491403AC0BF58F1C6

View on Hacker News

Recent submissions

Comment · 6 years ago
Comment · 6 years ago
Do you have any data to back this up? I should hope people would be embarrassed to post wildly unsubstantiated nonsense under a username linked to them.
Comment · 6 years ago
It's a random individual's approximation of the value added by their status as a government agency; and no, we patently /don't/ pay for it through taxes (unless you count subsidized borrowing, but that wasn't the original claim and it's more than a little nitpick-y).
Comment · 6 years ago
Comment · 6 years ago
So, I think what you're missing here is that the 'verification' is a formal specification about the behavior of the kernel, along with a set of proofs about that specification. This is vastly simplified (there are multiple levels of specification, last I checked), but that's the general idea. It's possible to have your compiler automatically derive the specification from your code (and, indeed, the seL4 people have written a Rust-like language called Cogent that does this), but formal verification inherently involves an understanding of what your code is /intended/ to do, and proving that it /does/. The compiler only has information about what it /does/, so it can only generate a specification that describes this. Information about your high-level intentions for the code has to be concretized somewhere, while ensuring that the spec conforms to it-- that's the job of the proof engineers.
Comment · 6 years ago
...why? If you don't like it, just don't install it. Not sure why you feel the need to ruin it for everyone else.
Comment · 6 years ago
Comment · 6 years ago
Comment · 6 years ago
I use a weechat relay over websockets (via nginx). It's a pretty convenient solution to the whole "my corporate firewall thinks I'm part of a botnet" issue (in addition to giving you channel history persistence).
Comment · 6 years ago
> they put a Tor client in the default distribution

Something that Mozilla is also doing[0], albeit with significantly more attention paid to the privacy-minded patches to Firefox made by the Tor Project. The end goal is to mainline these patches, such that Firefox gains the same privacy properties as the former-- they're working with the Tor team on this. If I wanted to use Firefox to access the Tor network, I'd just start a client locally and have the browser proxy traffic through it. Tor Browser's anti-fingerprinting measures are half the reason it's useful.

It's half-baked "features" like this that turn me away from Brave; there's this feeling that it's marketed towards the sort of well-meaning crowd that won't know the difference between "Tor tabs" and Tor Browser. This isn't a bad thing; rather, it's the misleading material surrounding such features published by the developers. Just look at the relevant page[1]; there's just a vague mention near the end of the article about using Tor Browser if you need "leakproof privacy", and the rest of the page carefully tiptoes around the fact that this really only prevents websites from knowing your IP address (and likely not even that, considering the similar fingerprinting properties of normal tabs and Tor tabs).

Excerpt:

"Also, web destinations can no longer easily identify or track a user arriving via Brave’s Private Tabs with Tor by means of their IP address."

To non-technical users, this reads more like "Brave (with Tor tabs) prevents websites from knowing who you are". It's lawyer-y, if that makes sense-- the whole thing just rubs me the wrong way. This is, of course, assuming said users even read the linked post; most will probably see 'Tor' and draw the relevant conclusions.

> even if the browser does not implement some of the anti-fingerprinting measures that Tor browser does

It's laughably easy to fingerprint Brave browser users, even compared to Firefox and its rather basic fingerprinting protections (at the moment). Though this is likely because Brave has a far smaller userbase.

[0] https://wiki.mozilla.org/Security/Tor_Uplift

[1] https://brave.com/tor-tabs-beta/

Comment · 7 years ago
How are participants prevented from simply splitting up their stakes?
Comment · 7 years ago
Comment · 7 years ago

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton